Gemini MCP Server

@aliargun

代码仓库

基本信息

分类

AI 与智能体

许可证

MIT

传输方式

stdio

作者

aliargun

关于 Gemini MCP Server

MCP server implementation for Google's Gemini API

#gemini #mcp-server #google-api

Gemini MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "gemini": {
      "command": "npx",
      "args": [
        "-y",
        "github:aliargun/mcp-server-gemini"
      ],
      "env": {
        "GEMINI_API_KEY": "your_api_key_here"
      }
    }
  }
}

Gemini MCP Server 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表，不代表服务没有工具。请查阅项目文档了解支持的能力。

点击上方按钮,在临时沙箱中运行此服务器并实时获取工具列表。

概览

What is Gemini MCP Server?

Gemini MCP Server is a Model Context Protocol (MCP) server that integrates Google’s latest Gemini AI models—including Gemini 2.5 with thinking capabilities—into any MCP-compatible development environment (e.g., Claude Desktop, Cursor, Windsurf). It provides tools for text generation, image analysis, token counting, model listing, embeddings, and self-documentation.

How to use Gemini MCP Server?

Configure your MCP client with the npx command npx -y github:aliargun/mcp-server-gemini and set the GEMINI_API_KEY environment variable. Once connected, use natural language in your client to invoke Gemini capabilities (e.g., “Use Gemini to review this code” or “Analyze this image with Gemini”).

Key features of Gemini MCP Server

Six built-in tools: text generation, image analysis, token counting, model listing, embeddings, and help
Supports latest Gemini 2.5 models with thinking, JSON mode, and Google Search grounding
Full MCP protocol via stdio communication for seamless client integration
Self-documenting help system accessible from within the editor
Modern TypeScript and ESM implementation
Configurable system instructions and conversation memory

Use cases of Gemini MCP Server

Complex reasoning tasks using Gemini 2.5 Pro’s 2M token context
Code review and analysis with adjustable temperature and model selection
Image analysis (vision) within an editor workflow
Research with grounding to incorporate Google Search results
Token counting and model listing for capacity planning

常见问题

What models are supported?

The server supports Gemini 2.5 Pro (2M tokens, thinking, JSON, grounding), Gemini 2.5 Flash (1M tokens, thinking, JSON, grounding), Gemini 2.5 Flash Lite (1M tokens, thinking, JSON), Gemini 2.0 Flash (1M tokens, JSON, grounding), and Gemini 1.5 Pro (2M tokens, JSON).

How do I get and set up an API key?

Obtain an API key from Google AI Studio (makersuite.google.com/app/apikey). Set it as the environment variable `GEMINI_API_KEY` in your MCP client configuration. Never commit the key to version control.

Which MCP clients are supported?

The server works with any MCP-compatible client that supports stdio transport, including Claude Desktop, Cursor, and Windsurf. Configuration examples for each are provided in the README.

What transport does the server use?

The server uses stdio transport, communicating via standard input/output with the MCP client. No HTTP or WebSocket transport is mentioned.

Are there any known limitations?

The README does not list specific limitations, but notes that API keys must be kept secure, and common issues can be resolved by checking client logs and internet connectivity. Troubleshooting guidance is available in the dedicated docs.

AI 与智能体分类下的更多 MCP 服务器

Rokha

Rokha AI LLC

Search a registry of agent skills and MCP servers, run them for real (sandbox or remote MCP, no install, a trace on every run), and hire proven agents for USDC.

fhirHydrant

faulkj

Open-source Node.js FHIR MCP server with SMART Backend Services, metadata-aware search/CRUD tools, compact responses, FHIRPath filtering, safe pagination, audit events, and terminology lookup.

strac-mcp-dlp

strac-io

Detect and redact PII, PHI, PCI and secrets — SSNs, credit cards, passports, API keys and cloud credentials — in text and in local files, using the Strac DLP API. Five tools: redact_text, detect_sensitive_data, detect_fi

Poly-Glot AI Workspace

hmoses

1,000+ fully localized prompt templates across 40+ categories. Compare Mode — send the same prompt to up to 9 AI providers (ChatGPT, Claude, Gemini, Grok, Groq, Copilot, Mistral, Perplexity, DuckDuckGo AI) and compare an

Legion MCP